Anchored VWAP

The volume-weighted average price accumulated from a specific starting bar you choose — an event, a swing point, or a session — rather than an automatic daily reset.

What it is

Anchored VWAP (AVWAP) is the volume-weighted average price calculated cumulatively from a specific bar that you choose — the 'anchor' — instead of resetting automatically at each session open the way standard VWAP does. Popularized by Brian Shannon, it answers a precise, descriptive question: from this exact starting point (an earnings release, a swing high or low, a breakout bar, a year open), what average price has been paid, weighting every bar's typical price by its volume? Because higher-volume bars pull the average harder, AVWAP is often described as tracking the average position cost of participants since the anchor event. The calculation itself is exact and deterministic; the honesty caveat is the anchor. The anchor is a human choice, usually made after the event has already proven significant — which is hindsight. An AVWAP anchored to 'the major low' looks meaningful precisely because you already know it was the major low. The line describes past transactions from a chosen origin; it does not predict where price will go, no anchor choice changes that, and this is educational content only — trading involves risk of loss.

How it works

From the anchor bar forward, keep two running sums: cumulative (typical price × volume) and cumulative volume. AVWAP on any bar is the first sum divided by the second: AVWAP = Σ(Pᵢ·Vᵢ) / Σ(Vᵢ), summed from the anchor bar to the current bar, where Pᵢ is usually hlc3 (high+low+close)/3. Nothing resets after the anchor — every bar since the anchor stays in the sums forever, which is what distinguishes it from rolling averages like VWMA (fixed window) and from session VWAP (daily reset). Two properties follow directly from the math. First, the line stiffens over time: as cumulative volume grows, each new bar is a smaller fraction of the total, so an AVWAP anchored months ago barely moves day to day while a fresh anchor swings with every bar. Second, implementation matters: the naive Pine approach of calling ta.vwap, or resetting sums on a condition that can fire more than once, silently produces a line that resets when you didn't intend it to. The correct implementation initializes both sums exactly once — on the first bar at or after the anchor time — and accumulates on every bar after, as in the example below.

How traders read it

• Above/below: price trading above an AVWAP means the current price is higher than the volume-weighted average paid since the anchor — a description of who is ahead on average since that event, not a directional forecast.
• Slope and distance: a rising AVWAP with price holding above it describes persistent buying at progressively higher volume-weighted prices since the anchor; large distance from the line describes an extended move relative to that average.
• Reaction zones: some traders watch how price behaves on a return to an AVWAP from a widely watched anchor (a major low, an earnings bar), on the logic that the average participant since that event breaks even there — a hypothesis about behaviour, not a guarantee of support or resistance.
• Age matters: a fresh AVWAP is responsive and easily crossed; an old one is heavily smoothed by accumulated volume and moves very slowly. Interpret crossings accordingly.
• Multiple anchors: comparing AVWAPs from several distinct events (e.g. two swing lows) shows how average transaction prices from different starting points converge or diverge.
• Always ask the hindsight question: would you have chosen this anchor before you knew how price played out? If not, the clean-looking history is partly survivorship.

Common settings

There is no length parameter — the only settings are the anchor and the price source. Common anchors: session, week, month, quarter, or year opens; earnings or news bars; prominent swing highs and lows; IPO or listing bars. The source is typically hlc3, though some implementations use close. On TradingView, an interactive time input (input.time with confirm = true) lets you click the anchor bar directly on the chart. Note that intraday volume differs across data feeds, so the same anchor can produce slightly different AVWAP values on different platforms.

Strengths

• Exact and deterministic once the anchor is fixed: same data, same anchor, same line — easy to verify and to reproduce across platforms.
• Volume weighting reflects where transactions actually concentrated, unlike time-only averages that treat a dead bar and a heavy bar equally.
• Ties analysis to a specific, stated event rather than an arbitrary rolling window, which makes the reasoning behind the line explicit and reviewable.
• No lookback-length parameter to curve-fit — the honesty burden moves to the anchor choice, where it is at least visible.
• Flexible across timeframes and instruments: the same cumulative math applies to an intraday breakout anchor or a multi-year low.

Pitfalls to watch

• Anchor hindsight bias: anchors are usually placed on bars already known to be important, so historical AVWAPs look more meaningful than any anchor you could have chosen in real time.
• Naive implementations reset wrong: using ta.vwap (which is session-anchored) or a reset condition that can re-trigger produces a silently incorrect line; the sums must initialize exactly once at the anchor.
• Increasing inertia: because volume accumulates forever, old AVWAPs respond very little to new information — a stale anchor can sit far from anything relevant to current trading.
• Feed dependence: volume differs across brokers and data vendors (especially in decentralized forex, where 'volume' is tick count), so AVWAP values are feed-specific, not universal levels.
• Support/resistance is a hypothesis, not a property: price crosses AVWAPs routinely; the 'average participant breakeven' story is an interpretation, not a mechanism that holds price.
• No predictive power: AVWAP is an exact summary of past transactions from a chosen bar — it forecasts nothing, and no anchor choice removes the risk of loss.

Pine v6 example

//@version=6
indicator("Anchored VWAP Example", overlay = true)

anchorTime = input.time(timestamp("2026-01-02 00:00 +0000"), "Anchor", confirm = true)
src = input.source(hlc3, "Source")

var float cumPV  = na
var float cumVol = na

bool atAnchor = time >= anchorTime and (na(time[1]) or time[1] < anchorTime)

if atAnchor
    // Initialize the sums exactly once, on the first bar at/after the anchor.
    cumPV  := src * volume
    cumVol := volume
else if time >= anchorTime
    cumPV  := cumPV + src * volume
    cumVol := cumVol + volume

float avwap = not na(cumVol) and cumVol > 0 ? cumPV / cumVol : na

plot(avwap, "Anchored VWAP", color = color.blue, linewidth = 2)

Pro tip: Stress-test your anchors for hindsight: before studying how price 'respected' an AVWAP from a major low, also anchor a few lines to unremarkable bars from the same period and compare. If arbitrary anchors produce similar-looking touches, the significance you saw was partly pattern-matching on a line that had to be near price anyway. Used honestly — anchored to events you would have identified in real time, with the feed-dependence of volume in mind — AVWAP is a precise descriptive reference. It is educational context, not a signal: no anchor predicts price, and trading involves risk of loss.

Educational only — not financial advice, not a recommendation to trade. No indicator is predictive; trading involves substantial risk of loss.